SCT-19749; No of Pages 11

Surface & Coatings Technology xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Surface & Coatings Technology

journal homepage: www.elsevier.com/locate/surfcoat

Characterization of soft magnetic spinel ferrite coating prepared by

plasma spray
Yi Liu, Shicheng Wei , Haoliang Tian, Hui Tong, Binshi Xu
National Key Laboratory for Remanufacturing, Academy of Armored Forces Engineering, Beijing 100072, China

a r t i c l e

i n f o

Article history:
Received 9 April 2014
Accepted in revised form 12 September 2014
Available online xxxx
Keywords:
Plasma spray
Spinel ferrite
Tribology

a b s t r a c t
The primary objective of this work is to investigate the microstructure and tribological performance of magnetic
nickelzincmagnesium ferrite coating prepared through plasma spray. The phase composition, morphology and
valence state of the ferrite coating are studied by X-ray diffraction (XRD), scanning electron microscopy (SEM)
and X-ray photoelectron spectroscopy (XPS). It is found that the ferrite coating shows a spinel structure. The
as-sprayed coating exhibits nanocrystalline with the grain size of 100200 nm. According to the XPS analysis,
the ferrite coating contains both Fe2+ and Fe3+, a few Ni ions are reduced to metallic Ni. Tribological performance
of the ferrite coating is tested in dry friction and 3.5 wt.% NaCl solution using a ball-on-disk test. The wear loss of
the ferrite coating is lower than the substrate in both dry friction and NaCl solution under the same wear test conditions. The wear mechanism of the ferrite coating is dominated by delamination and brittle fracture. Microwave
absorbing property of the ferrite shows a high value of magnetic loss factor in the frequency range of 112 GHz.
The results suggest that plasma spray is a promising technology for the production of magnetic ferrite coatings.
2014 Elsevier B.V. All rights reserved.

1. Introduction
The majority applications of thermal spray are in the eld of protective coatings and the main functions of the overlay coating are to protect
the underlying substrate from heat, wear or corrosion [1]. New opportunities are now emerging in advanced functional materials such as magnetic materials, catalytic materials and bioactive materials. Thermal
spray provides advantages in the efciency for the preparation of such
materials. In recent years, thermal spray deposition of microwave absorbing materials has attracted attention because of the relative higher
adhesion strength and thinner thickness of the coating compared with
the traditional polymer composite method. Among the few examples,
Lisjak et al. and Begard et al. focused on the atmosphere plasma spray
to prepare barium hexagonal ferrites as electromagnetic absorbers [2,
3]. Other researchers also made correlation studies in this eld. Bobzin
et al. prepared SrFe12O19 coatings through atmospheric plasma spray
using two kinds of feedstock powders (type A and type B). The feedstock of type A consisted of spray-dried spherical agglomerates of
micrometric SrFe12O19 particles while the feedstock of type B was synthesized by reactively sintered agglomerates of SrCO3 and Fe2O3. They
found that type B agglomerates exhibited magnetic properties, which
were promising for electromagnetic wave absorption applications [4].
Furthermore, Zhao et al. investigated the microwave absorption properties and the complex permittivity of Fe/FeAl2O4 coatings deposited by
reactive plasma spray using Al/Fe2O3 powders. They found that the
Corresponding author. Tel.: +86 10 66718541; fax: +86 10 66717144.
E-mail address: wsc33333@163.com (S. Wei).

real part and the imaginary part of the complex permittivity increased
with Al concentration in the frequency range of 8.212.4 GHz [5].
Most of these papers focused on the preparation procedure and give
trials for thermal spray in the application of microwave absorption
technology.
Nickelzincmagnesium (NZM) spinel ferrites are soft magnetic
materials and can be used for radar absorbing materials (RAMs) which
are mostly applied in the frequency range of 118 GHz [68]. The
NZM ferrites have various forms in products, such as sheet, paints,
lms and powders. Apart from these products, the ferrite can be
blended with organic binders to produce microwave absorbing coatings. Recently much effort has been made to improve the performance
of RAMs and simplify the preparation procedure [9]. Plasma spray is a
promising technique to meet this need. If NZM ferrites can be fabricated
by plasma spray with proper magnetic properties, the preparation efciency can be improved. However, the deposition of NZM ferrite coating
through plasma spray has not been reported so far.
RAMs usually serve in various environments. The RAMs particularly
coated on the ships or tanks is often partly scratched by the obstacle in
the environment. It can be attributed to the mechanical damage. In addition, the RAMs also can be scratched by the ying glass when a bomb
went off [10]. Surface damage generated by abrasive contact will limit
the life of RAMs and reduce their reliability. Therefore, tribology performance is also necessary for the RAMs. Unfortunately, the studies on the
tribological behavior of RAMs are scarce up to now. The aim of this work
is to prepare NZM ferrite coating through plasma spray. Small amounts
of lanthanum are added to the NZM ferrite in order to improve the magnetic property and rene the crystal size [1113]. Tribological tests are

http://dx.doi.org/10.1016/j.surfcoat.2014.09.029
0257-8972/ 2014 Elsevier B.V. All rights reserved.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

Table 1
Plasma spray parameters of the coating.
Voltage
(V)

Current
(A)

Ar
(m3/h)

H2
(m3/h)

N2
(m3/h)

Spray distance
(mm)

Powder feeding rate

(g/min)

120

380

3.2

0.4

0.6

100

30

conducted in dry friction and 3.5 wt.% NaCl solution considering that
equipments coated with RAMs sometimes serve in a marine environment. The effect of environmental conditions on the tribological behavior is discussed.
2. Material and methods
2.1. Powder production
The feedstock powder consisted of spray-dried spherical agglomerates of micrometric spinel ferrite particles with a composition of
Ni0.5Zn0.4Mg0.1La0.05Fe1.95O4. Stoichiometric amounts of MgO, ZnO,
NiO, La2O3 and Fe2O3 as starting materials were mixed and homogenized in a ball mill. The particles were synthesized by a solid state reaction. These starting materials were sintered at 1300 C for 2 h to achieve
reactive sintering, so as to make the original reagents convert to a spinel
phase. During this process, the size of reagents got bigger and

distributed inhomogeneously. In order to achieve ne products, the

sintering products were broken down to powders again. The powders
were heat-treated at 900 C for 1 h after the shattering process, which
is used to reduce the internal stress and defects. At last, the powders
were cooled down to room temperature naturally. However, these
powders were not suitable for plasma spray due to their poor
owability. In order to prepare spherical agglomerates and improve
their owability, spray-drying process was selected to produce plasma
spray materials. The preparation procedure was as follows: the ferrite
powders were used as raw materials and polyvinyl alcohol as the binder. The particles were agglomerated into spherical granules, using a
spray-drier attached to a cyclone; the entry and exit temperatures
were 240 C and 140 C, respectively. The spray dried powders were
sieved between 200 mesh and 400 mesh to separate the bigger and
ner ones, so that the size of agglomerates fell within the scope of plasma spray. The size distribution of the particles nally used in plasma
spray was in the range of 200 mesh to 400 mesh.

(b)

(a)

(c)

(d)
1.8x10-5

0.0090
0.0085
0.0080
0.0075
0.0070
0.0065
0.0060
0.0055
0.0050
0.0045
0.0040
0.0035
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005

1.6x10-5
1.4x10-5
1.2x10-5
1.0x10-5

Substrate

8.0x10-6
6.0x10-6
4.0x10-6
2.0x10-6

50

100

150

200

250

300

350

400

450

500

0.0
550

0.0080
0.0075
0.0070
0.0065
0.0060
0.0055
0.0050
0.0045
0.0040
0.0035
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005

1.4x10-5
1.2x10-5
1.0x10-5
8.0x10-6
6.0x10-6
4.0x10-6

Coating

2.0x10-6

50

100

150

200

250

300

350

400

450

500

Fig. 1. SEM image of the ferrite coating: (a) Cross-section morphology, (b) image of the interface, (c) CTE of the substrate and (d) CTE of the coating.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

0.0
550

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

2.2. Coating deposition

Mild steel substrate was used in this study. The substrate was
ground with abrasive papers up to 600 grit, cleaned with acetone, and
then sandblasted using brown alumina grit prior to plasma spray. The
ferrite coatings were prepared through the high efciency supersonic
plasma spraying equipment developed by National Key Laboratory for
Remanufacturing. The equipment consumed lower energy and gas
ow rate [14]. The torch used a single anode Laval nozzle. The anode
nozzle internal diameter was 4.5 mm. Mechanical compression was
the primary mechanism of the torch. It achieved a high reliability and
heavy energy density [15]. The powders were injected through radial
position and the internal diameter of the injector was 3 mm. Detailed
operating parameters were listed in Table 1. The nitrogen was used as
a powder carrier gas. The carrier gas ow rate was 0.6 m3/h. It was determined by the powder feeding rate and plasma jet power.
2.3. Characterization

3. Results and discussion

3.1. Morphological study
Fig. 1a shows the typical cross-section image of the ferrite coating. It
can be seen that smooth and dense coatings are fabricated by plasma

(311)

(511)

(444)

coating
20

(422)

powder

(440)

LaFeO3

(400)

Tribological tests were carried out in a reciprocating mode using a

ball-on-disk sliding wear testing apparatus (CETR UMT-3). The counter
part was a Si3N4 ball. The diameter of the counterpart was 4 mm and the
overall test duration time was 30 min. The friction tests were carried out
at room temperature with a frequency of 10 Hz, a normal load of 10 N
and a wear track length of 4 mm. The wet friction test was performed
in 3.5 wt.% NaCl solution. The wear volume was determined using a
3D non-contact surface mapping proler (LEXT OLS4000, Olympus). It
should be noted that the surface state was one of the key factors to be
considered. The as-sprayed coatings exhibited a signicantly different
surface roughness. In order to decrease the friction coefcient, the surface roughness should be as low as possible [16]. Therefore, before tribological tests, the samples were ground using SiC papers (up to
1000 mesh) and then polished with diamond slurries until the average
surface roughness in the same order of magnitude.

spray. The coating is constituted of well attened splats and partial molten particles. The partial molten particles are marked by the arrow. During the process of plasma spray, a large amount of microparticles are
melted in the high-temperature ames (more than 104 K). These molten droplets strike the substrate surface and form a micron sized lamellar structure under the action of high-speed ame. There is an obvious
interface between the coating and the substrate. The magnication

(220)

2.4. Tribological test procedure

Fig. 2. TEM image of the ferrite coating.

Intensity (a.u.)

The phase structures of the ferrite powder and coating were studied
by D8 advance X-ray diffraction (XRD) with = 0.154 nm Cu-K radiation. The cross sectional samples were prepared by metallographic
cutting, hot-mounting in resin, grinding with abrasive papers (up to
2000 mesh) and polishing with a diamond slurry. The prepared sample
was etched by the Nital. The morphology patterns of powder and coating samples were observed using a eld emission scanning electron microscopy (FEI, Navo NanoSEM 450) coupled with an energy dispersive
spectra (EDS, OXFORD Feature Max). The microstructure of the ferrite
nanocrystalline was studied by a transmission electron microscope
(TEM, JEOL JEM 2000FX). The valence states of elements in ferrite
coating were analyzed by X-ray photoelectron spectroscopy (XPS, VG
Scientic ESCALAB 220i-XL, USA). The thermal expansion coefcients
of both substrate and ferrite coating were measured from 30 C to
500 C using a NETSCH Dilatometer Model DIL 402C. The complex
permeability and complex permittivity were measured in the range
of 118 GHz by an HP8722ES network analyzer. For this purpose,
the ferrite powders were homogeneously dispersed into wax matrix
and compacted into rings for the permeability and permittivity measurements. The size of the ring was 7 mm in outer diameter, 3 mm in
inner diameter and 2 mm in thickness. The ferritewax composites
contained 80 wt.% of ferrite.

30

40

50

60

70

80

2 Theta (degree)
Fig. 3. XRD patterns of the ferrite samples, one phase is the spinel phase, and the second
phase is the LaFeO3.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

image of the interface is showed in Fig. 1b. The interface shows a loose
connection and is accompanied by some unmelted particles (Fig. 1b, circle) and cracks (Fig. 1b, arrow). It is attributed to excessive shrinkage
rates of the large number of molten particles deposited on the cold
metal substrate at a rapid cooling rate. As is known to all, a lamellar
structure is built up after molten particles impact, atten, adhere, and
solidify on the target substrate [17]. The CTE of the substrate and the ferrite coating was shown in Fig. 1c and d, respectively. It can be seen that
the CTE of ferrite coating is much smaller than the substrate. Thus it results in the mismatch of thermal stress. Because of the high cooling rate
and mismatch of thermal stress between the deposit coating and the
steel substrate, defects such as pores and microcracks easily develop
at the interface of the two different materials, which may degrade the
mechanical property [18]. Moreover the wetting property maybe
inuence by the poor connection between the coating and the substrate
and inappropriate contact angles of the molten particles lead to
the porosity between the coating and substrate. Therefore, suitable
wettabilitytemperature relation between the molten particles and
the cold substrate is required to produce good quality coatings [19]. In
this paper, the bond layer is not used for the connection of the metal
substrate and ceramic coating, which may also result in a loose connection between the two heterogeneous material interfaces. Fig. 2 shows
the TEM image of the as-sprayed ferrite coating. A number of ferrite
grain boundaries can be obviously seen. The crystals exhibit an equiaxial
rather than a columnar shape because the sample is cut from the axial
position of coating surface. The observed grains of the ferrite coatings
are nanocrystalline (as marked by an arrow). The grain size is in the

(a)

3.2. XRD analysis

Fig. 3 shows the XRD patterns of sintered ferrite powder and the ferrite coating. The phase analysis shows that the sintered powder samples
are well crystalline. The sharp and strong diffraction peak around (311)
plane provides a clear evidence for the formation of the cubic spinel
structure of the ferrite as reported everywhere [21]. The several intense
peaks correspond to planes of (220), (311), (400), (422), (511) and
(440). Furthermore, the diffraction peaks of the ferrite coating become
broad as compared with the powder. There are two mechanisms for
the XRD peak broadening. Firstly, the background noise and broadness
of the peaks are characteristic of particles with nanometer dimensions.
It happens in the nano-sized particles because there are insufcient diffraction centers that cause the line broadening [22]. This is conrmed by
the TEM image. Secondly, residual stress and crystal defects in thermal
spray coatings also have an effect on XRD peak broadening and signal/
background ratio, particularly when only a proportion of particles are
melted [23,24]. In addition, the XRD peaks of the ferrite coating are

(b)

40000

Fe 2p3/2

50000

Fe

Ni

Ni 2p3/2

Fe 2p1/2

35000

48000

30000

Intensity (cps)

Intensity (cps)

range of 100200 nm. The white parts in the TEM picture are also the
ferrite grains. They show different contrasts with the black grains. The
formation of the nanocrystal can be considered as the result of the fast
solidication process. Plasma spray is a rapid process; the residence
time of the ferrite powders inside the plasma jet is less than 10 3 s.
Therefore, there is little time for the molten particles to grow, which results in the formation of nano-crystals [20]. In addition, micro-structures may also be present in the coating.

25000
20000

46000
Ni 2p1/2

44000
42000

15000
40000

10000
740

730

720

710

875

700

870

(c)

860

855

850

845

(d)

64000

68000

Zn

62000

Mg

66000

60000

64000

Intensity (cps)

Intensity (cps)

865

Binding Energy (eV)

Binding Energy (eV)

58000
56000
54000
52000

62000
60000
58000
56000
54000

50000

52000

48000
1050

1045

1040

1035

1030

1025

Binding Energy (eV)

1020

1015

50000

1308

1306

1304

1302

1300

1298

1296

Binding Energy (eV)

Fig. 4. XPS spectra of the ferrite coating: (a) Fe, (b) Ni, (c) Zn and (d) Mg. (For interpretation of the references to color in this gure, the reader is referred to the web version of this article.)

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

3.3. XPS analysis

Fig. 4 represents XPS core level spectra of Fe 2p, Mg 1s, Ni 2p and Zn
2p of the ferrite coating sample. The Fe 2p binding energy is shown in
Fig. 4a. It is reported that Fe 2p photoelectron peaks of oxidized iron
are associated with satellite peaks. The analysis of Fe 2p3/2 photoelectron peak is complicated due to a high temperature spray process and
ion distribution in the ferrite system. Therefore, the satellite peaks of
Fe3+ and Fe2+ are important to distinguish their chemical states. It is localized on 6 eV and 8 eV for FeO and Fe2O3 above their basic

3
-2

14
12
10
8
6
4
2
0

Test sample
Fig. 6. Variations in the volume loss of the samples.

photoelectron peak [29]. According to the previous reports [30], the Fe

2p3/2 binding energy of FeO and Fe2O3 is in the range of 709.3
709.7 eV and 710.4711.2 eV, respectively. As shown in the gure, the
black line is the measured spectrum. The as-synthesized ferrite coating
sample shows a main peak at 710.1 eV. A satellite peak is observed at
716.2 eV, which is about 6 eV above the main peak. Therefore, it can
be concluded that Fe2+ ions exist in the ferrite coating. However, the
binding energy of the main peak locates between FeO and Fe2O3. Thus,
the ferrite coating contains both its trivalent state and divalent state.
The deconvoluted Fe 2p peak is shown in the color solid lines. The red
line stands for the total tting chemical states of iron. The photoelectron
peak at 709.3 eV in the pink line stands for Fe2+ in FeO while the photoelectron peak at 711.2 eV in the green line stands for Fe3+ in Fe2O3.
The blue line stands for the Fe2+ satellite spectrum. Besides, a peak locates in the range of 720725 eV, which corresponds to Fe 2p1/2 photoelectron peak.
The XPS core level spectrum of Ni is shown in Fig. 4b. The black solid
line is the measured curve. The red line is the tting results of all the different chemical states of Ni. The binding energy of Ni 2p3/2 is associated
with three peaks, which correspond to the Ni3+ in Ni2O3, Ni2+ in NiO
and metallic Ni, respectively. The peak value at 852.4 eV in the green
line corresponds to the metallic Ni while the peak value at 854.7 eV in
the pink solid line corresponds to the Ni2+ in NiO [31]. In addition, the
peak value at 855.9 eV in the blue line corresponds to the Ni3 + in
Ni2O3. Another peak is observed at the binding energy of 872.3 eV,

(b)
0.9

1.0

0.8

0.9

0.7
0.6

substrate
dry
3.5% NaCl

0.5

Friction coefficient

Friction coefficient

(a)

substrate in dry friction

substrate in solution
coating in dry friction
coating in solution

16

Volume loss (10 mm )

clearly shifted to lower angles compared with the ferrite powder. It is

due to zinc loss during the plasma spray process. Yan et al. conducted
extensive fundamental studies to improve the understanding of the
zinc loss on different sizes of splats. They determined XRD spectra
from different sizes of splats and also found Bragg peaks exhibit a shift
to lower 2 with increased size of splats [25]. Moreover, a little amount
of the secondary phase (LaFeO3) is observed in both the sintered ferrite
and the coating. The secondary phase indicates that lanthanum does not
form a solid solution with spinel ferrite or it has a limited solid solubility
in the lattice. The ionic radii of metal ions depend on their coordination
number. The ionic radii of the studied ions such as Ni2+ (0.69 ), Zn2+
(0.74 ), Fe2+ (0.74 ) and Fe3+ (0.64 ) are similar. However, the ionic
radius of La3+ (1.06 ) is much larger than the other ions. Lanthanum
has a very low solubility in spinel lattice due to its bigger ionic radii.
The amount of La3+ ions is limited when they are used to substitute
for Fe3 + ions in spinel lattice, as the substitution of La exceeds the
solid solubility, the redundant La3+ ions aggregate on the grain boundaries and produce the secondary phase. During structure renement, it
is found that the occupancy of the octahedral site tends to increase for
La-doped compositions, which allows few of La3 + ions to enter into
the lattice [26]. Lanthanum is a non-magnetic element; the substitution
of La can modify the magnetic properties of the ferrite. On the other
hand, the redundant La forms a secondary phase that segregates in
grain boundaries, which prevents the grain growth and renes the crystal size. In many cases on thermal sprayed ferrites, the breaking of the
ferrite structure is reported, which usually happens in the hexagonal
ferrite materials [27]. The required time for the crystallization process
of hexagonal ferrite which involves various peritectic reactions is too
long compared to the very short solidication process. A few hexaferrite
phases can be retained in the coatings [4]. In this research, spinel ferrite
instead of hexagonal ferrite is selected. The spinel phase is a quick formation process and it is the rst phase developed from the liquid [28].
Therefore the spinel phase is retained and no structure breaking of the
coating is observed.

0.8
0.7

coating
dry
3.5% NaCl

0.6

0.4

0.5
0.3

300

600

900

1200

1500

1800

300

Time/s

600

900

Time/s

Fig. 5. Friction coefcients of the samples: (a) substrate and (b) coating.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

1200

1500

1800

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

which is typical of Ni 2p1/2 in NiO. According to the handbook of XPS

[32], the peak of Ni2+ in NiO is observed at 853.8 eV. However, the
same peak appears at 854.7 eV in the ferrite coating. The binding energy
of photoelectron peak not only depends on the chemical state, but also
relies on its molecular environment. The increase in the binding energy
of Ni2+ may be due to the cation distribution in the spinel structure.
Nickel prefers the octahedral sites and it occupies the octahedral site
during the ferrite growth. Moreover, there is an interesting phenomenon in the XPS spectrum. The metallic Ni is discovered, which indicates
that as the ferrite powder is sprayed on the substrate during plasma
spray, some Ni ions are reduced to metallic Ni. It may be related to the
reductive atmosphere of the Ar/H2 plasma and the carrier gas
(nitrogen). This phenomenon is also observed in other reports. Tao prepared NiFe2O4 through molten salt synthesis in KCl ux using Fe2O3 and
NiO. The powders containing 1.2 wt.% polyvinyl alcohols were sintered
in air and nitrogen atmospheres. Metallic nickel was detected in the
samples that sintered in nitrogen atmosphere [33]. Furthermore, Yan
et al. also discovered a reductive process for metal ions. The oxidation
state of Fe changed from Fe3+ to Fe2+ during the spray process [25].
Fig. 4c shows the XPS core level spectra of Zn 2p. Zn2+ ions prefer
the tetrahedral site in the spinel structure. According to literature [34],
there is some difference for Zn2 + in the nanocrystalline materials. If
Zn2 + ions occupy octahedral sites, the binding energy of Zn 2p peak
will be slightly higher. The binding energy of Zn 2p3/2 in the tetrahedral
site is seen at 1021.4 eV while the binding energy of Zn 2p3/2 in the octahedral site is seen at 1023.2 eV. In this work, Zn 2p3/2 peak is observed
at 1021.5 eV, which indicates that Zn2+ ions occupy the tetrahedral site
in the coating. Another peak at 1044.7 eV stands for Zn 2p1/2.
Mg2+ ions usually prefer the octahedral sites in oxygen lattice. Thus,
in ferrites, when cations distribute in two different sites, two XPS peaks
or a peak broadening of the photoelectron peak is expected. However, in
this research, there is only one peak for Mg 1s. The binding energy is
1303.9 eV (Fig. 4d), which corresponds to the Mg2+. The result shows
that Mg2+ ions distribute in the octahedral site.
3.4. Wear behavior of the samples
The friction coefcient is an important parameter to evaluate the
tribology property [35]. Fig. 5 shows the variations of friction coefcient of both the substrate and the ferrite coating with a frequency of
10 Hz and normal load of 10 N. Initially, the friction coefcient of the

substrate (Fig. 5a) increases rapidly until it reaches a peak value, and
then a gradual decrease is followed to a steady state value. The
steady friction coefcient is about 0.82 in dry friction. However, the
friction coefcient of the substrate was lower in 3.5 wt.% NaCl solution than that in dry friction and the steady value is 0.75. With regard
to the ferrite coating (Fig. 5b), friction coefcient suddenly increases
from 0.5 to the peak value in a very short time and then shows a
gradual decrease to the end of the test. The friction coefcient of
the ferrite coating nally exhibits a steady value of 0.88 in dry condition and 0.76 in solution, respectively. The decrease of friction coefcient is due to the lubrication action of the solution. The evolution
of friction coefcient exhibits two stages in the tribology test. The
rst stage is the running-in period. There is a rise in the friction coefcient plot, which represents the formation of wear debris. The next
stage is the steady wear state. A steady and lower friction coefcient
is measured. The lower friction coefcient can be attributed to that
debris is either trapped in the wear track and roll under the indent
or pushed out of the ball path [36]. The volume wear loss of the substrate and the coating in dry friction and NaCl solution is showed in
Fig. 6. The volume loss of the substrate is larger than the ferrite coating in both dry friction and the solution under the same condition. As
the substrate sliding in NaCl solution, the volume wear loss is larger
than in dry friction. The steel substrate exhibits poor corrosion resistance in NaCl solution. Accordingly, the corrosive wear aggravates
the volume wear loss of the substrate. However, the result of the ferrite coating is contrary to the substrate. The volume loss of the coating in solution is lower than in dry friction. Both the ferrite coating
and the counterpart belong to ceramic materials, which have better
corrosion resistance in NaCl solution. The addition of solution plays
a role in a lubricant, which reduces the volume loss of the coating.
Fig. 7a shows the worn surface of the substrate in dry friction. Huge
material accumulation and apparent plow scratches can be found
around the wear track. It is accepted that during the unlubricated sliding
wear process, oxidation reaction occurs rst in the real contact area due
to the synthetic effect of contact stress and frictional heat. Fine debris
generates and detaches from the oxidation regions under repeated cyclic stress [37]. It is seen that the worn surface of the substrate shows
signs of abrasive wear and adhesion wear when sliding against a Si3N4
ball. However, the worn surface of the substrate in 3.5 wt.% NaCl solution exhibits a severe damage. A large number of apparent abrasive
grooves and pits (generated by the delamination of the materials) are

Fig. 7. Worn surface of the substrate: (a) dry friction and (b) wet friction.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

identied on the worn surface (Fig. 7b). In summary, the wear scar of
the substrate immersing in 3.5 wt.% NaCl solution is associated with
abrasion, generation of cracks, material and debris removal. The morphology analysis of wear scar is in accordance with the uctuations of
friction coefcient. Furthermore, the transparent solution subsequently
changes into a red turbid one, which suggests that corrosion plays an
important role in the wear test. The wear mechanisms are abrasive
wear and corrosive wear. The corrosion reactions of the substrate in
3.5 wt.% NaCl solution include the following steps [38]:

0:5O2 H2 O 2e2OH

Fe OH Cl FeClOH e

H2 O

FeClOH FeOH Cl e
FeOH H Fe

Fe

2OH FeOH2

FeOH2 0:5O2 H2 O eFeOH3 OH :

As the steel dissolves, the corrosion products play the role of abrasive particles. These particles are taken away from the worn surface by
solution. The fresh substrate is continually exposed to the corrosive
media, which speeds up the corrosion process. Compared with dry
friction, the substrate is severely damaged and shows large wear
scratches. It is notable that the lower friction coefcient is not equal to
lower wear volume loss in the wet tribology test. The number of corrosion pits and cracks continually develops due to corrosive wear. The
worn surface of the sample gets rough under corrosive attack, which reduces the energy of abrasive wear and accelerates the wear damage.
In order to understand better the coating damage induced by the tribology test, a polished surface of the plasma spray coating before the tribology test is shown in Fig. 8a. It can be seen that the coating surface is

Fig. 8. Worn surface of the ferrite coating: (a) polished coating surface before the tribology test, (b) worn surface of the coating in dry friction, (c) worn surface of the coating in wet friction
and (d) EDS analysis of the selected area.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

porous. The worn surface of the ferrite coating under a normal load
of 10 N in dry friction is shown in Fig. 8b. Extensive brittle fracture
occurs, leading to splat detachment on the worn surface in dry friction. The coating material is pressed and distorted under the normal
load. This process increases the coating stress, decreases the adhesive strength, aggravates the coating deformation and produces a
large number of debris. The worn surface of the ferrite coating in
wet friction is shown in Fig. 8c. SEM micrograph exhibits a large
amount of pits. Small chip can be found on the surface. The wear
mechanism is brittle spalling. According to the previous research
[39], friction can induce the initiation of cracks between two splats
of the same lamella. The cracks propagate through the splat boundary under cyclic load, the solution quickly lls in the cracks, which
accelerates the crack growth rate and results in the nal exfoliation
on the worn surface. During the friction test in NaCl solution, hydrodynamic effects such as enhanced cooling of the surface and ushing
of the debris also play a role of reducing the friction coefcient and
volume wear loss. The thicker corrosion product and the bigger pits

lled with liquid, the better lubrication and less contact area can be
provided [40]. EDS of the scar (Fig. 8d) shows that the worn surface
consists of Ni, Zn, Mg, La, Fe, O, Si, Na and Cl.
The wear debris is accumulated after the tribology test. Fig. 9a
shows the SEM image of substrate debris in the dry friction test.
The debris is primarily in the form of akes, which suggests it is continually attened and deformed by the counterpart. In order to reveal the distinct variations of chemical composition, the wear
debris is investigated by EDS. The debris is composed of O, Si and
Fe. The mole fraction of oxygen is 57.4 at.%, which indicates that
the debris is heavily oxidized. However, the morphology of the substrate debris in NaCl solution is quite different from that in dry friction (see Fig. 9b). The debris is surrounded by the NaCl particles.
The composition is Fe, Cl, Si, Na and O. The silicon comes from the
counterpart (Si3N4 ball). Fig. 9(c) shows the wear debris of the ferrite
coating in dry friction. The debris also exhibits a aky structure.
Composition analysis shows that it is composed of Ni, Si, O and Fe.
The Fe, O and Si are the major composition. Among these elements,

Fig. 9. SEM images of wear debris: (a) substrate in dry friction, (b) substrate in NaCl solution and (c) ferrite coating in dry friction.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

Fe and O come from the ferrite coating while Si comes from the
counterpart. It has to be noted that the oxygen can also be a result
from the oxidation during the sliding wear. Unfortunately, the debris
of the ferrite coating in NaCl solution is too few that it is hard to collect accurately.
The worn surface of the counterpart (Si3N4 ball) is also investigated. Fig. 10a shows the counter surface of the substrate in dry
friction. A large number of apparent parallel plows exhibit in the
picture. According to EDS analysis, the worn surface is composed
of Si and Fe. The presence of Fe suggests the adhesion of the substrate. The SEM image of the counterpart against the substrate in
NaCl solution is shown in Fig. 10b. It can be seen that the surfaces
of a Si3N4 ball show obvious grooves and adhesion slice. EDS analysis reveals that the worn surface is composed of Si, N, O and Fe.
The presence of Fe and O also conrms the adhesion regime on
the worn surface. Interestingly, the worn surface of the counterpart
has a higher content of Fe when it slides against the substrate in
NaCl solution. This result corresponds to severe adhesion of the
substrate in wet friction. The detection of Fe on the counterpart

suggests a continuous tribo-lm layer on the silicon nitride ball.

The worn surface of the counterpart ball against ferrite coating in
dry friction and solution is shown in Fig. 10c and d. Apart from
some shallow plows on the surface of the counterpart ball in dry
friction, the wear track of the counterpart shows a smooth surface.
Little debris is present on the worn surface. The debris most likely
comes from the ferrite coating.
3.5. Microwave absorbing property
Fig. 11(a) and (b) displays a frequency dependence of the complex
permeability and complex permittivity of the ferrite, respectively. It
can be seen the real part of the permeability () decreases rapidly
with increasing frequency until it reach the minimum value at the frequency of 7.8 GHz. Then it increases with increasing frequency. The
imaginary part of permeability () decreases with the increasing frequency below 13 GHz and then it is almost constant. As shown in
Fig. 11(b), both the real part () and the imaginary part () of permittivity are found almost constant within the entire frequency range. The

Fig. 10. Worn surface of the counterpart ball sliding against: (a) substrate in dry friction, (b) substrate in NaCl solution, (c) ferrite coating in dry friction and (d) ferrite in NaCl solution.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

10

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

(a)

(b)
2.4

'
''

2.1

1.8

1.2

1.5

'

'

0.9

'
''

3
2

''

0.6
0.3

''

0.0
2

10

12

-1

14

16

-2

18

Electromagnetic loss factor

(c)

10

12

14

16

18

Frequency/GHz

Frequency/GHz
1.4
1.2

tan

1.0

tan

0.8
0.6
0.4
0.2
0.0
2

10

12

14

16

18

Frequency/GHz
Fig. 11. Frequency dependence of the ferrite powders: (a) complex permeability, (b) complex permittivity and (c) electromagnetic loss factors.

value of is in the range of 5.55.9 and the value of is about 0.05

0.25. It is well known that microwave absorbing property can be described by electromagnetic loss factor ( tan ). It is calculated from
the formulas [41]:
tan

tan

:
0

The electromagnetic loss factor is illustrated in Fig. 11c. As can be

seen, the magnetic loss factor shows a very high value in the frequency
range of 112 GHz and the peak value is about 1.21 at the frequency of
5.25 GHz. Moreover the magnetic loss factor is much higher than the dielectric loss factor when the frequency is below 13 GHz; however they
are almost same in the frequency of 1318 GHz. The dielectric loss factor
is very low in the investigated frequency. The peak value of tan is just
0.04. The result indicates that magnetic loss plays a dominant role in the
absorbing mechanism.
4. Conclusions
NZM spinel ferrite coating is successfully synthesized by plasma
spray using the spray dried ferrite powders. The coating exhibits a
cubic spinel structure with a secondary phase of LaFeO3. The diffraction
peaks of the ferrite coating become broader as compared with the powders. Cross section image shows an obvious interface between the

coating and the substrate. TEM morphology analysis indicates that the
ferrite coating is composed of nanoparticles. The average size of these
nanoparticles is in the range of 100200 nm. During plasma spray, a reductive process is discovered for Ni ions. Some Ni ions are reduced to
metallic Ni due to the reductive atmosphere. Tribology behavior of the
substrate and the ferrite coating sliding against Si3N4 counterpart ball
is investigated in dry friction and NaCl solution. The friction coefcient
of both the substrate and coating samples is lower in solution. The substrate suffers adhesion wear and abrasive wear. The ferrite coating
shows a better wear resistance than the substrate. The wear process of
the coating is dominated by delamination and brittle fracture. Microwave absorbing property of the ferrite shows a high value of magnetic
loss factor in the frequency range of 112 GHz and the peak value is
about 1.21 at the frequency of 5.25 GHz.
Acknowledgments
This paper is nancially supported by the National Natural Science
Foundation of China (No. 51222510) and 973 Project (2011CB013403).
References
[1] Sanjay Sampath, J. Therm. Spray Technol. 19 (2010) 921949.
[2] D. Lisjak, K. Bobzin, K. Richardt, Marion Bgard, Giovanni Bolelli, Luca Lusvarghi,
et al., J. Eur. Ceram. Soc. 29 (2009) 23332341.
[3] M. Begard, K. Bobzin, G. Bolelli, A. Hujanen, P. Lintunen, D. Lisjak, et al., Surf. Coat.
Technol. 203 (2009) 33123319.
[4] K. Bobzin, G. Bolelli, M. Bruehl, A. Hujanen, P. Lintunen, D. Lisjak, et al., J. Eur. Ceram.
Soc. 31 (2011) 14391449.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029

Y. Liu et al. / Surface & Coatings Technology xxx (2014) xxxxxx

[5] D. Zhao, F. Luo, W.C. Zhou, D.M. Zhu, Surf. Coat. Technol. 205 (2011) 42544259.
[6] C.H. Peng, C.C. Hwang, J. Wan, J.S. Tsai, S.Y. Chen, Mater. Sci. Eng. B 117 (2005)
2736.
[7] A.R. Bueno, M.L. Gregori, M.C.S. Nbrega, J. Magn. Magn. Mater. 320 (2008)
864870.
[8] N. Singh, A. Agarwal, S. Sanghi, Paramjeet Singh, Physica B 406 (2011) 687692.
[9] H.L. Fan, W. Yang, Z.M. Chao, Compos. Sci. Technol. 67 (2007) 34723479.
[10] L. Yang, Y.F. Wang, Y.S. Zhang, Y.Z. Lv, New Technol. New Process 1 (2009) 97100
(in Chinese).
[11] Ankush Thakur, R.R. Singh, P.B. Barman, J. Magn. Magn. Mater. 326 (2013) 3540.
[12] N. Chen, K. Yang, M.Y. Gu, J. Alloys Compd. 490 (2010) 609612.
[13] L.W. Deng, L. Ding, K.S. Zhou, S.X. Huang, Z.W. Hu, B.C. Yang, J. Magn. Magn. Mater.
323 (2011) 18951898.
[14] B.S. Xu, H.J. Wang, S. Zhu, X.B. Liang, Manuf. Technol. Mach. Tool 2 (2003) 3033 (in
Chinese).
[15] H.J. Wang, B.S. Xu, P. Zhang, Z.P. Wang, C.S. Zhai, The spray gun of ultrasonic plasma
spray, Patent, 01101077.0 (China).
[16] G. Darut, H. Ageorges, A. Denoirjean, P. Fauchais, Surf. Coat. Technol. 217 (2013)
172180.
[17] S.Y. Tao, Z.J. Yin, X.M. Zhou, C.X. Ding, Tribol. Int. 43 (2010) 6975.
[18] H.Y. Zhu, Y.R. Niu, C.C. Lin, L.P. Huang, H. Ji, X.B. Zheng, Ceram. Int. 39 (2013)
101110.
[19] R. Comesaa, F. Quintero, F. Lusquios, M.J. Pascual, M. Boutinguiza, A. Durn, et al.,
Acta Biomater. 6 (2010) 953961.
[20] Y. Zhu, M. Huang, J. Huang, et al., J. Therm. Spray Technol. 8 (1999) 219222.
[21] T.H. Ting, R.P. Yu, Y.N. Jau, Mater. Chem. Phys. 126 (2011) 364368.
[22] M.A. Gabal, Abdullah M. Asiri, Y.M. AlAngari, Ceram. Int. 37 (2011) 26252630.

11

[23] T.C. Totemeler, R.N. Wright, W.D. Swank, Metall. Mater. Trans. A 34 (2003)
22232231.
[24] X.B. Liu, B. Zhang, J. Mater. Sci. 37 (2002) 32293239.
[25] Q. Yan, R.J. Gambino, S. Sampath, L.H. Lewis, L. Li, E. Baumberger, et al., Acta Mater.
52 (2004) 33473353.
[26] P.K. Roy, J. Bera, Mater. Res. Bull. 42 (2007) 7783.
[27] K. Bobzin, T. Schlaefer, M. Begard, et al., Surf. Coat. Technol. 205 (2010) 10151020.
[28] N. Langhof, D. Seifert, M. Gbbels, J. Tpfer, J. Solid State Chem. 182 (2009)
24092416.
[29] V.K. Mittal, P. Chandramohan, Santanu Bera, M.P. Srinivasan, S. Velmurugan, S.V.
Narasimhan, Solid State Commun. 137 (2006) 610.
[30] Marina V. Bukhtiyarova, Aleksandra S. Lvanova, Elena M. Slavinskaya, Lyudmila M.
Plyasova, Vladimir A. Rogov, Vasily V. Kaichev, et al., Fuel 90 (2011) 12451256.
[31] V.V. Kaichev, A.Yu. Gladky, I.P. Prosvirin, A.A. Saraev, M. Hvecker, A. Knop-Gericke,
et al., Surf. Sci. 609 (2013) 113118.
[32] John F. Moulder, F. Stickle William, Peter E. Sobol, D. Bomben Kenneth, Handbook of
X-ray Photoelectron Spectroscopy, Perkin-Elmer cooperation, Eden Prairie, 1992.
[33] Y.Q. Tao, Z.Y. Li, K.C. Zhou, Ceram. Int. 39 (2013) 865869.
[34] L. Zhang, Y. Wang, Q.Q. Ni, Mater. Chem. Phys. 124 (2010) 10291033.
[35] J.B. Cheng, X.B. Liang, Z.H. Wang, B.S. Xu, Tribol. Int. 60 (2013) 140146.
[36] P. Henty, M.-J. Pac, C. Rousselot, M.-H. Tuilier, Surf. Coat. Technol. 223 (2013) 7986.
[37] C. Guo, J.S. Zhou, J.M. Chen, J.R. Zhao, Y.J. Yu, H.D. Zhou, Wear 270 (2011) 492498.
[38] Q. Bao, D. Zhang, D.D. Lv, P. Wang, Corros. Sci. 65 (2012) 405413.
[39] P.P. Psyllaki, M. Jeandin, D.I. Pantelis, Mater. Lett. 47 (2001) 7782.
[40] H.B. Lee, D.S. Wuu, C.Y. Lee, C.S. Lin, Tribol. Int. 44 (2011) 16031609.
[41] X.G. Huang, J. Zhang, L.X. Wang, et al., J. Alloys Compd. 540 (2012) 137140.

Please cite this article as: Y. Liu, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.09.029